Novel Application of a Multistate Model to Evaluate the Opioid Use Disorder Care Cascade: A Retrospective Cohort Study

BackgroundEvaluating the opioid use disorder (OUD) care cascade can improve OUD treatment retention and care. ObjectivesTo identify risk and protective factors for retention among patients in OUD treatment. MethodsWe conducted a retrospective cohort study among patients diagnosed with OUD using data from the Rhode Island (RI) All-Payer Claims Database from 2011 to 2019. Patients who initiated treatment (Stage 2) were classified into sub-stages of retention (Stage 3) corresponding to multistate modeling states capturing early retention (sub-stage 1), short and long-term retention (sub-stage 2), and short and long-term disengagement (sub-stage 3). The association of baseline characteristics with state transitions were evaluated. ResultsA cohort of 6,939 RI residents diagnosed with OUD included 41% aged 40 to 60 years, 57.6% male, and 70.8% Medicaid beneficiaries. In sub-stage 1, cannabis (Relative risk ratios (RRR) = 1.16; 95% confidence interval (CI) = 1.04,1.29) and cocaine use disorders (RRR=1.15; 95% CI=1.05,1.25) increased early disengagement risk after engagement. Medicaid beneficiaries were less likely to experience early disengagement (RRR=0.81; 95% CI =0.76,0.87). In sub-stage 2, alcohol (RRR=1.29; 95% CI=1.13,1.47) or cocaine use disorders (RRR=1.18; 95% CI=1.01,1.40) increased risk of disengagement among patients in the retention states. In sub-stage 3, tobacco (RRR=1.10; 95% CI=1.01,1.21) and alcohol (RRR=1.14; 95% CI=1.03,1.27) use disorders were associated with re-engagement from disengaged states. ConclusionThe multistate model applied to a cohort of patients initiating medication for OUD led to the identification of factors associated with treatment engagement and retention. These results may guide strategies to sustain treatment among OUD patients.